People chronically underestimate how long tasks will take. In their original article, the present authors (M. M. Roy, N. J. S. Christenfeld, & C. R. M. McKenzie) suggested a simple, broadly applicable explanation: Biased predictions result from biased memories. In their comment article, D. Griffin and R. Buehler suggested that in many domains in which this memory-bias account appears to outpredict their own account, theirs actually makes no prediction at all. However, the present authors did not suggest that only 1 theory is right but that theirs is consistent with data that prior theories, including their own, cannot explain. Ignoring memories of past tasks is not a complete explanation for the phenomenon if the memories people could consult are themselves biased. Nonetheless, underestimating future task duration is almost certainly multiply determined, and thus our account and theirs can coexist.